Heterogeneous Integration of Graphene and HfO₂ Memristors

Abstract

The past decade has seen a growing trend toward utilizing (quasi) van der Waals growth for the heterogeneous integration of various materials for advanced electronics. In this work, pulsed‐laser deposition is used to grow HfO2 thin films on graphene/SiO2/Si. As graphene is easily damaged under standard oxide‐film deposition conditions, the process needs to be adjusted to minimize the oxidation and the collision‐induced damage. A systematic study is conducted in order to identify the crucial deposition parameters for diminishing the defect concentration in the graphene interlayer. For evaluating the quality of graphene, it is mainly relied on data obtained from Raman spectroscopy, using approaches beyond the Tuinstra‐Koenig relation. The results show that the defects are mainly a consequence of the high kinetic energy of the plasma‐plume particles. Using a relatively high Ar process pressure, a sufficiently low defect concentration is ensured, without compromising the quality of the HfO2 thin film. This enabled us to successfully prepare memristive devices with a filamentary type of switching, utilizing the graphene layer as a bottom electrode. The findings of this study can be easily transferred to other systems for the development of oxide electronic devices.

Programmable Retention Characteristics in MoS2-Based Atomristors for Neuromorphic and Reservoir Computing Systems

“…This approach offers a solution to the degradation issues related to the CBRAM. In addition, 2D material-based memristors have a high potential for several applications owing to their high ON/OFF ratio, threshold switching characteristics with a low-compliance current, increased switching speed, ultralow power consumption (fJ per switching), and gigahertz operation. − In terms of scalability, the atomic-scale thickness of 2D materials enables their applications in a variety of structures, such as three-dimensional (3D) vertical or lateral channels, thus presenting the possibility of increasing device density. − Specifically, research using a 2D electrode edge contact selector has demonstrated the capability to reduce off currents, thus improving power consumption and highlighting the potential for advances in energy-efficient neuromorphic computing research. , Furthermore, recent 2D memristor studies suggest the emulation of various synaptic properties and their suitability for neuromorphic computing as well as structural improvements. Table lists the previous studies related to STM and LTM characteristics and synaptic properties of other 2D memristors. ,,,− Here, we demonstrate the coexistence of STM and LTM properties in MoS 2 -based atomristors.…”